How to monitor both train and validation metrics at the same step?

🀗Transformers
22

@Mariam Did you find the solution of displaying both training and validation loss at the same time?

23

@sgugger there is missing validation loss and eval loss while running run_qa.py. Configuration are:
!python run_qa.py
–model_name_or_path deepset/roberta-base-squad2
–output_dir /content/squad
–do_train
–do_eval
–overwrite_output_dir
–per_device_train_batch_size 16
–per_device_eval_batch_size 16
–train_file /content/drive/MyDrive/qa_model/transformer_qa/temp_data/train.json
–validation_file /content/drive/MyDrive/qa_model/transformer_qa/temp_data/valid.json
–num_train_epochs 30
–logging_dir /content/logs
–evaluation_strategy epoch
–logging_steps 10
–eval_steps 10 --gradient_accumulation_steps 16
–max_train_samples 80 --max_eval_samples 32
–save_total_limit 2 --run_name bert-base-high-lr --learning_rate 3e-5 --warmup_steps 100 --weight_decay 0.03 --logging_strategy epoch